Martin Kroeker [Wed, 22 Feb 2017 09:31:39 +0000 (10:31 +0100)]
Merge pull request #1104 from martin-frbg/lapack-comma
LAPACK: fix missing comma on continued lines
Martin Kroeker [Wed, 22 Feb 2017 07:40:39 +0000 (08:40 +0100)]
Fix missing comma on continued line
EXTERNAL declaration of subroutines missed a comma before the continuation line,
causing a strange run-together name to appear in the object when compiled with ifort.
Martin Kroeker [Wed, 22 Feb 2017 07:39:06 +0000 (08:39 +0100)]
Fix missing comma on continued line
EXTERNAL declaration of subroutines missed a comma before the continuation line,
causing a strange run-together name to appear in the object when compiled with ifort.
Martin Kroeker [Wed, 22 Feb 2017 07:34:20 +0000 (08:34 +0100)]
Fix missing comma on continued line
EXTERNAL declaration of subroutines missed a comma before the continuation line,
causing a strange run-together name to appear in the object when compiled with ifort.
Martin Kroeker [Wed, 22 Feb 2017 07:32:20 +0000 (08:32 +0100)]
Fix missing comma in continued line
EXTERNAL declaration of subroutines missed a comma before the continuation line,
causing a strange run-together name to appear in the object when compiled with ifort.
Martin Kroeker [Tue, 21 Feb 2017 21:58:30 +0000 (22:58 +0100)]
Merge pull request #1103 from vladimir-ch/fix-lapacke-ormbr
LAPACKE: fix wrong matrix size in ?ormbr
Vladimir Chalupecky [Tue, 21 Feb 2017 20:57:18 +0000 (21:57 +0100)]
LAPACKE: fix wrong matrix size in ?ormbr
Changes made upstream in Reference LAPACK in
https://github.com/Reference-LAPACK/lapack/pull/128
Martin Kroeker [Tue, 21 Feb 2017 14:26:14 +0000 (15:26 +0100)]
Merge pull request #1098 from martin-frbg/amodra-power8
Power8 inline assembly fixes
Martin Kroeker [Tue, 21 Feb 2017 14:19:56 +0000 (15:19 +0100)]
Merge pull request #1101 from martin-frbg/martin-frbg-patch-1
LAPACKE: fix wrong number of columns in ?ormlq
Martin Kroeker [Tue, 21 Feb 2017 07:26:39 +0000 (08:26 +0100)]
Merge pull request #1102 from brada4/develop
Correct Apollo Lake CPUID identification in dynamic_arch builds
Andrew [Mon, 20 Feb 2017 22:54:59 +0000 (23:54 +0100)]
detect apollo lake for real
Martin Kroeker [Mon, 20 Feb 2017 15:20:43 +0000 (16:20 +0100)]
LAPACKE: fix wrong number of columns in ?ormlq
Copied from lapack https://github.com/Reference-LAPACK/lapack/pull/127 by vladimir-ch (with earlier changes from echeresh's
PR 115 "lapacke_*ormlq_work: move declarations under if" there as they touched some of the same files)
Martin Kroeker [Fri, 17 Feb 2017 09:30:09 +0000 (10:30 +0100)]
Merge pull request #1091 from staticfloat/sf/corei5_7600k
CPUID mappings for Core i5-7600K (Kaby Lake)
Martin Kroeker [Mon, 13 Feb 2017 22:38:50 +0000 (23:38 +0100)]
Power8 inline assembly fixes
Quoting patch author amodra from #1078
Lots of issues here.
- The vsx regs weren't listed as clobbered.
- Poor choice of vsx regs, which along with the lack of clobbers led to
trashing v0..v21 and fr14..fr23. Ideally you'd let gcc choose all
temp vsx regs, but asms currently have a limit of 30 i/o parms.
- Other regs were clobbered unnecessarily, seemingly in an attempt to
clobber inputs, with gcc-7 complaining about the clobber of r2.
(Changed inputs should be also listed as outputs or as an i/o.)
- "r" constraint used instead of "b" for gprs used in insns where the
r0 encoding means zero rather than r0.
- There were unused asm inputs too.
- All memory was clobbered rather than hooking up memory outputs with
proper memory constraints, and that and the lack of proper memory
input constraints meant the asms needed to be volatile and their
containing function noinline.
- Some parameters were being passed unnecessarily via memory.
- When a copy of a
Martin Kroeker [Sun, 12 Feb 2017 16:00:17 +0000 (17:00 +0100)]
Merge pull request #1096 from martin-frbg/pkg-config
Build only openblas.pc for pkg-config and install it from cmake as well
Martin Kroeker [Sun, 12 Feb 2017 13:38:32 +0000 (14:38 +0100)]
Add cmake template for openblas.pc
Martin Kroeker [Sun, 12 Feb 2017 13:37:33 +0000 (14:37 +0100)]
Create and install openblas.pc in cmake builds
Martin Kroeker [Sun, 12 Feb 2017 13:35:48 +0000 (14:35 +0100)]
Create and install only a single openblas.pc file
Martin Kroeker [Sun, 12 Feb 2017 13:34:03 +0000 (14:34 +0100)]
Rename blas.pc.in to openblas.pc.in
Martin Kroeker [Sun, 12 Feb 2017 13:30:29 +0000 (14:30 +0100)]
Merge pull request #1095 from martin-frbg/lapack370-cmake
Update cmakefiles for netlib 3.7.0
Martin Kroeker [Sun, 12 Feb 2017 12:49:49 +0000 (13:49 +0100)]
Add zlasyf_aa to lapack.cmake
Martin Kroeker [Sun, 12 Feb 2017 00:59:30 +0000 (01:59 +0100)]
Add another bunch of lapack 3.7 functions to cmake list
Martin Kroeker [Sun, 12 Feb 2017 00:37:35 +0000 (01:37 +0100)]
Add LAPACK 3.7 files not mentioned in announcement
Martin Kroeker [Sat, 11 Feb 2017 23:40:16 +0000 (00:40 +0100)]
Update cmake file list for lapacke 3.7.0
Martin Kroeker [Sat, 11 Feb 2017 22:11:26 +0000 (23:11 +0100)]
Update cmake file list for lapack 3.7.0
Martin Kroeker [Sat, 11 Feb 2017 19:48:41 +0000 (20:48 +0100)]
Merge pull request #1094 from martin-frbg/cmake-1
Update cmakefiles with changes from netlib 3.6.1
Martin Kroeker [Sat, 11 Feb 2017 18:56:02 +0000 (19:56 +0100)]
Reflect name change of lapacke_mangling.h template
Martin Kroeker [Sat, 11 Feb 2017 18:54:02 +0000 (19:54 +0100)]
Add new functions from LAPACK 3.6.1
Martin Kroeker [Sat, 11 Feb 2017 16:41:39 +0000 (17:41 +0100)]
Merge pull request #1093 from martin-frbg/restore-cmakeinstall
Restore cmake install target
Martin Kroeker [Sat, 11 Feb 2017 15:43:46 +0000 (16:43 +0100)]
Add cmake install target
Add CMAKE install target (based on patch provided by PrimarchOfTheSpaceWolves in #957)
This was originally merged as 988 but accidentally reverted by my subsequent PR the following day
Elliot Saba [Fri, 10 Feb 2017 23:23:34 +0000 (15:23 -0800)]
Add `exfamily == 9` case (Kaby Lake) to dynamic arch detection
Elliot Saba [Fri, 10 Feb 2017 22:47:10 +0000 (14:47 -0800)]
CPUID mappings for Core i5-7600K (Kaby Lake)
Martin Kroeker [Wed, 8 Feb 2017 00:01:18 +0000 (01:01 +0100)]
Merge pull request #1084 from isuruf/develop
Install pkg-config files
Martin Kroeker [Wed, 8 Feb 2017 00:00:32 +0000 (01:00 +0100)]
Merge pull request #1087 from grisuthedragon/enable-a12
Enable EXCAVATOR kernels for A12-9800
Martin Koehler [Tue, 7 Feb 2017 20:38:28 +0000 (21:38 +0100)]
Enable EXCAVATOR kernels for A12-9800
Martin Kroeker [Tue, 7 Feb 2017 10:40:41 +0000 (11:40 +0100)]
Merge pull request #1085 from vladimir-ch/lapacke_laswp_work
LAPACKE: fix incorrect value of lda_t in lapacke_?laswp_work
Vladimir Chalupecky [Tue, 7 Feb 2017 08:21:46 +0000 (09:21 +0100)]
LAPACKE: fix incorrect value of lda_t in lapacke_?laswp_work
Fixed in Reference LAPACK in commit:
https://github.com/Reference-LAPACK/lapack/pull/118/commits/
07e1fbd89752bed74d35c48e92287d467646a158
Isuru Fernando [Mon, 6 Feb 2017 06:29:48 +0000 (11:59 +0530)]
Install pkg-config files
Martin Kroeker [Sat, 4 Feb 2017 16:25:43 +0000 (17:25 +0100)]
Merge pull request #1076 from ashwinyes/develop_20170130_thunderx2t99
More optimized implementations for ThunderX2T99
Ashwin Sekhar T K [Thu, 2 Feb 2017 06:10:35 +0000 (22:10 -0800)]
THUNDERX2T99: Add optimized S/D/C/Z COPY Implementations
Ashwin Sekhar T K [Wed, 1 Feb 2017 07:25:41 +0000 (23:25 -0800)]
THUDNERX2T99: Add optimized D/C/Z ASUM Implementations
Ashwin Sekhar T K [Tue, 31 Jan 2017 06:10:45 +0000 (11:40 +0530)]
LAPACK: Fix lapack-test errors in ARM64 threaded version
Ashwin Sekhar T K [Fri, 27 Jan 2017 09:26:00 +0000 (01:26 -0800)]
THUNDERX2T99: Add optimized CASUM Implementation
Ashwin Sekhar T K [Mon, 30 Jan 2017 06:39:04 +0000 (12:09 +0530)]
THUNDERX2T99: Rename labels in for DDOT and SNRM2
Ashwin Sekhar T K [Fri, 27 Jan 2017 09:11:58 +0000 (01:11 -0800)]
THUNDERX2T99: Remove Duplicate Code
Ashwin Sekhar T K [Wed, 25 Jan 2017 11:14:59 +0000 (03:14 -0800)]
THUNDERX2T99: Add Optimized CGEMM Implementation
Zhang Xianyi [Wed, 25 Jan 2017 14:17:05 +0000 (22:17 +0800)]
Merge pull request #1074 from ashwinyes/develop_20170116_thunderx2t99_sgemm
Add more THUNDERX2T99 Optimized APIs
Ashwin Sekhar T K [Wed, 25 Jan 2017 07:14:09 +0000 (23:14 -0800)]
Update .gitignore
Ashwin Sekhar T K [Wed, 25 Jan 2017 07:13:47 +0000 (23:13 -0800)]
Benchmark: Add MFlops print in iamax benchmark
Ashwin Sekhar T K [Wed, 25 Jan 2017 04:50:23 +0000 (20:50 -0800)]
Benchmarks: Avoid building lapack benchmarks when NO_LAPACK=1
Ashwin Sekhar T K [Tue, 24 Jan 2017 16:09:29 +0000 (21:39 +0530)]
THUNDERX2T99: Add threaded SNRM2 Implementation
Ashwin Sekhar T K [Tue, 24 Jan 2017 09:19:49 +0000 (14:49 +0530)]
ARM64: Rename kernel files to have consistent naming
Ashwin Sekhar T K [Thu, 19 Jan 2017 10:27:13 +0000 (15:57 +0530)]
THUNDERX2T99: Add Optimized CNRM2 Implementation
Ashwin Sekhar T K [Thu, 19 Jan 2017 08:57:02 +0000 (00:57 -0800)]
THUNDERX2T99: Add Optimized SNRM2 Implementation
Ashwin Sekhar T K [Wed, 18 Jan 2017 08:39:04 +0000 (00:39 -0800)]
Update .gitignore
Ashwin Sekhar T K [Thu, 19 Jan 2017 05:26:17 +0000 (10:56 +0530)]
THUNDERX2T99: Add threaded DDOT Implementation
Ashwin Sekhar T K [Thu, 19 Jan 2017 05:23:48 +0000 (10:53 +0530)]
THUNDERX2T99: Add Optimized DDOT Implementation
Ashwin Sekhar T K [Wed, 18 Jan 2017 08:57:11 +0000 (00:57 -0800)]
THUNDERX2T99: Improve SGEMM
Ashwin Sekhar T K [Tue, 17 Jan 2017 07:16:23 +0000 (23:16 -0800)]
THUNDERX2T99: Improve DGEMM
Ashwin Sekhar T K [Tue, 17 Jan 2017 08:28:54 +0000 (00:28 -0800)]
THUNDERX2T99: Add Optimized DAXPY Implementation
Zhang Xianyi [Tue, 17 Jan 2017 14:02:21 +0000 (22:02 +0800)]
Merge pull request #1068 from sva-img/develop
Added MSA optimised rot functions.
Shivraj Patil [Tue, 17 Jan 2017 06:45:07 +0000 (12:15 +0530)]
Added rot functions.
Signed-off-by: Shivraj Patil <shivraj.patil@imgtec.com>
Ashwin Sekhar T K [Wed, 11 Jan 2017 09:37:11 +0000 (15:07 +0530)]
THUNDERX2T99: Add Optimized SGEMM Implementation
Ashwin Sekhar T K [Wed, 11 Jan 2017 07:47:10 +0000 (13:17 +0530)]
ARM64: Let target VULCAN inherit THUNDERX2T99 properties
Martin Kroeker [Mon, 16 Jan 2017 15:03:53 +0000 (16:03 +0100)]
Merge pull request #1067 from martin-frbg/msysinst
Fix DESTDIR support for cygwin/msys2 install
Martin Kroeker [Mon, 16 Jan 2017 14:15:46 +0000 (15:15 +0100)]
Fix DESTDIR support for cygwin/msys2 install
fixes #1066
Zhang Xianyi [Mon, 16 Jan 2017 05:20:10 +0000 (13:20 +0800)]
Merge pull request #1061 from ashwinyes/develop_aarch64_vulcan_thunderx_patch
Add new targets for ARM64
Martin Kroeker [Wed, 11 Jan 2017 16:40:06 +0000 (17:40 +0100)]
Update Makefile.install (#1064)
* Update Makefile.install to reflect name change of lapacke_mangling.h source
Werner Saar [Wed, 11 Jan 2017 11:37:45 +0000 (12:37 +0100)]
Merge pull request #1063 from wernsaar/develop
prepared kernel/setparam-ref.c for UNROLL values, that are not a power of two
Werner Saar [Wed, 11 Jan 2017 10:56:50 +0000 (11:56 +0100)]
prepared kernel/setparam-ref.c for UNROLL values, that are not a power of two
Werner Saar [Wed, 11 Jan 2017 09:30:46 +0000 (10:30 +0100)]
Merge pull request #1062 from wernsaar/develop
prepared parameter.c for UNROLL values, that are not a power of two
Werner Saar [Wed, 11 Jan 2017 08:50:28 +0000 (09:50 +0100)]
prepared parameter.c for UNROLL values, that are not a power of two
Werner Saar [Wed, 11 Jan 2017 06:29:17 +0000 (07:29 +0100)]
prepared lapack/lauum for UNROLL values, that are not a power of two
Ashwin Sekhar T K [Tue, 10 Jan 2017 08:55:55 +0000 (14:25 +0530)]
ARM64: Add Cavium THUNDERX2T99 Target
Ashwin Sekhar T K [Tue, 10 Jan 2017 07:23:47 +0000 (12:53 +0530)]
ARM64: Fix auto detect of ARM64 cpus
Andrew Pinski [Fri, 17 Jul 2015 04:08:03 +0000 (00:08 -0400)]
THUNDERX: Add optimized version of daxpy
This is better for single core but does not change anything for multiple cores
Martin Kroeker [Tue, 10 Jan 2017 18:09:49 +0000 (19:09 +0100)]
Merge pull request #1060 from martin-frbg/lapacke-mingw
Split LAPACKE 3.7.0 obj list (take 2, missed splitting the actual ar command invocation)
Martin Kroeker [Tue, 10 Jan 2017 16:11:35 +0000 (17:11 +0100)]
Split LAPACKE 3.7.0 obj list (take 2)
Missed the splitting of the actual ar call
Werner Saar [Tue, 10 Jan 2017 15:00:28 +0000 (16:00 +0100)]
Merge pull request #1059 from wernsaar/develop
updated some level1 funcions, that are not thread save
Werner Saar [Tue, 10 Jan 2017 13:05:07 +0000 (14:05 +0100)]
updated some level1 funcions, that are not thread save
Werner Saar [Tue, 10 Jan 2017 10:30:08 +0000 (11:30 +0100)]
Merge pull request #1058 from wernsaar/develop
prepared lapack/potrf functions for UNROLL values, that are not a pow…
Werner Saar [Tue, 10 Jan 2017 09:50:28 +0000 (10:50 +0100)]
prepared lapack/potrf functions for UNROLL values, that are not a power of two
Andrew Pinski [Thu, 16 Jul 2015 07:30:16 +0000 (03:30 -0400)]
THUNDERX: Add an optimized version of ddot
Andrew Pinski [Tue, 10 Jan 2017 06:27:36 +0000 (11:57 +0530)]
ARM64: Add Cavium THUNDERX Target
Ashwin Sekhar T K [Mon, 9 Jan 2017 13:18:39 +0000 (18:48 +0530)]
VULCAN: Add optimized DGEMM implementation
Ashwin Sekhar T K [Tue, 4 Oct 2016 08:50:20 +0000 (01:50 -0700)]
ARM64: Add the VULCAN Target
Ashwin Sekhar T K [Tue, 4 Oct 2016 08:24:28 +0000 (01:24 -0700)]
CORTEXA57: Add assembly kernels for copy routines
Zhang Xianyi [Tue, 10 Jan 2017 05:58:26 +0000 (13:58 +0800)]
Merge pull request #1055 from ksraste/develop
Add msa optimization for AXPY, COPY, SCALE, SWAP
jiahaipeng [Sun, 11 Dec 2016 09:09:50 +0000 (09:09 +0000)]
Adding multi-threading for copy, dot, rot, and asum funcitons
jiahaipeng [Sun, 11 Dec 2016 09:02:18 +0000 (09:02 +0000)]
modify the blas_l1_thread.c for support multi-threded for L1 fuction with return value
Martin Kroeker [Mon, 9 Jan 2017 19:45:26 +0000 (20:45 +0100)]
Merge pull request #1057 from martin-frbg/lapacke-mingw
Split the obj list of LAPACKE 3.7.0
Martin Kroeker [Mon, 9 Jan 2017 17:29:53 +0000 (18:29 +0100)]
Split the obj list of LAPACKE 3.7.0
Split obj list to allow building with mingw (argument list too long for the msys ar)
kaustubh [Mon, 9 Jan 2017 12:57:23 +0000 (18:27 +0530)]
Add msa optimization for AXPY, COPY, SCALE, SWAP
Signed-off-by: kaustubh <kaustubh.raste@imgtec.com>
kaustubh [Mon, 9 Jan 2017 12:52:09 +0000 (18:22 +0530)]
Add msa optimization for AXPY, COPY, SCALE, SWAP
Signed-off-by: kaustubh <kaustubh.raste@imgtec.com>
Werner Saar [Mon, 9 Jan 2017 12:38:56 +0000 (13:38 +0100)]
Merge pull request #1054 from wernsaar/develop
prepared lapack/getrf functions for UNROLL values, that are not a pow…
Werner Saar [Mon, 9 Jan 2017 11:57:26 +0000 (12:57 +0100)]
prepared lapack/getrf functions for UNROLL values, that are not a power of two
Zhang Xianyi [Mon, 9 Jan 2017 10:52:42 +0000 (05:52 -0500)]
Merge branch 'z13' into develop
Conflicts:
CONTRIBUTORS.md
Zhang Xianyi [Mon, 9 Jan 2017 10:48:09 +0000 (05:48 -0500)]
Add USE_TRMM=1 for IBM z13 in kernel/Makefile.L3
Werner Saar [Mon, 9 Jan 2017 10:17:38 +0000 (11:17 +0100)]
Merge pull request #1053 from wernsaar/develop
prepared driver/level3 functions for UNROLL values, that are not a po…
Werner Saar [Mon, 9 Jan 2017 09:38:15 +0000 (10:38 +0100)]
prepared driver/level3 functions for UNROLL values, that are not a power of two
projects / platform / upstream / openblas.git / log